Jul 2026· International Conference on Information Photonics· pp. 1-6· 0 citations· 27 references
Computer Science
TL;DR
WREN is an LLIE neural network with double U-Net-like structures trained end-to-end with a scale-invariant loss function, which gives robustness against the illumination scaling, and achieves the state-of-the-art performance across multiple datasets.
Abstract
This paper proposes a neural network for low light image enhancement (LLIE) based on retinex theory to make LLIE robust for various dynamic range scenes. The retinex theory is an image formulation model inspired by a human color perception hypothesis, where a low light image is decomposed into intrinsic color context (i.e., reflectance map) and scene-dependent illumination (i.e., illumination map). Due to non-uniqueness of its decomposition, existing retinex-based LLIE methods often fail to achieve stable decomposition, which lead to over-enhancement. Typically, they are sensitive to the dynamic ranges that vary in different lighting conditions. To tackle this issue, we propose WREN: An LLIE neural network with double U-Net-like structures. WREN consists of two U-Net-like sub-networks. The first network has one encoder and two decoders that decompose an input image into the reflectance and illumination maps. The second network with a customized Transformer block between an encoder and a decoder only enhances the illumination map obtained from the first network: This completely follows the assumption of the retinex theory. Finally, the enhanced illumination map is recombined with the reflectance map. The network is trained end-to-end with a scale-invariant loss function, which gives robustness against the illumination scaling. Numerical results show that our method achieves the state-of-the-art performance across multiple datasets. Our code is available online.
Low-light image enhancement has made significant progress through both traditional Retinex methods and deep learning techniques. Traditional Retinex-based methods decompose images into illumination and reflectance components to mimic human perception of brightness and color. However, these methods often struggle with noise suppression and detail preservation, particularly under severe low-light conditions. Recent Transformer-based methods, such as RetinexFormer and Restormer, have improved restoration performance by modeling long-range dependencies, but they still insufficiently explore the interaction between illumination variations and spatial--semantic features. To address these limitations, we propose Illumination-Aware Attention-based Transformer (IA2former), a novel low-light image enhancement model that explicitly models illumination-aware feature interactions. By integrating an Illumination-Aware Attention mechanism and an Illumination-Aware Loss function, IA2former effectively captures long-range dependencies, improves detail restoration, and preserves spatial structures under challenging illumination conditions. Experimental evaluations on the LOL-v1 and LOL-v2 datasets demonstrate that IA2former achieves a favorable overall balance across PSNR, SSIM, and LPIPS, obtaining the best performance on multiple metrics and remaining competitive on others. These results validate the effectiveness and robustness of the proposed illumination-aware modeling strategy for low-light image enhancement.
Tianqi Jiang· Poster Volume 0007 The 2026...· 0 citations
Low-light enhancement is a crucial task in computer vision; it can improve either the subjective experience of viewers or the usability of computer vision systems designed for normal-light images. In this paper, a variational Retinex model in the image domain is developed for low-light enhancement, which infuses classical/fractional differentiation of the input image into the illumination/reflectance component by means of a structure/texture-aware map (SAM/TAM). Firstly, the SAM (TAM) is generated by the inverse square of classical (fractional) differentiation of the input image. Secondly, the regularization term of illumination (reflectance) is defined by utilizing the SAM (TAM) as a weighted matrix, and an illumination guidance term is incorporated into the objective function. The illumination guidance term encourages the estimated illumination to encompass more structural information by penalizing deviation of illumination from the illumination pre-estimated by a dark channel prior to a guided image filtering. Finally, an alternative algorithm is employed to solve the minimization problem involved in the model. The performance of the proposed method is evaluated on three datasets for low-light enhancement and compared with eight state-of-the-art Retinex methods, qualitatively and quantitatively. Evaluation results show that the proposed method generally achieves higher performance in terms of low-light enhancement.
This work validates the design effectiveness of decoupling global and local representations within a frozen backbone, and establishes a new baseline for parameter-efficient enhancement.
Yanpeng Cao, Yue Wang, MingHui Liang et al.· Pattern Analysis and Applica...· 0 citations